Up-Scaling of Microbial Co-Cultures

Abstract

Microbial fermentations are used to produce pharmaceuticals, chemicals, and fuels. Side product formation along with considerable losses in productivities and yields are often encountered in transition from laboratory- to production-scale. This is mainly caused by inevitable heterogeneity of large, stirred reactors with aeration, especially in the commonly used fed-batch operation mode. Previous studies of this challenge have proposed either improving reactor homogeneity or increasing tolerance of the production host. Microbial co-cultures have also been suggested because they facilitate side-product utilization.

Purpose of this work was to examine the scale-up of microbial co-cultures. More specifically it was studied whether the adverse conditions caused by scale-up affect co-cultures less than the respective monocultures.

In this work monocultures of two bacterial species, Escherichia coli and Acinetobacter baylyi, and four co-cultures of them with and without gene deletions were cultivated in batch-mode in a 1 L aerated stirred reactor. Both batch and fed-batch operation modes were simulated with a model that incorporated a 70-compartment hydrodynamic representation of a 30 m3 reactor, population balances, and unstructured kinetics. The simulations were performed also with an ideal reactor model at the 1 L-scale.

Due to internal recycling of acetate, which inhibited E. coli but was the preferred substrate for A. baylyi, the co-cultures had 10 % to 50 % higher specific growth rates compared to the single strain cultivations. Simulations showed more modest improvements ranging from 3 % to 20 % in both studied scales. Specific growth rates and consequently biomass productivities were consistently smaller at the simulated large-scale with both mono- and co-cultures. Based on the simulations it seems that co-cultures do not scale-up any better than the respective monocultures, but appropriate scale-down experiments are required for verification. However, they do have the potential to alleviate side-product inhibition also at industrial scale.